revert and fix demo data

data-raw/loanbook_demo.R

@@ -1,148 +1,50 @@
 library(dplyr)
-library(r2dii.data)
-library(rlang)
-library(tidyr)
-
-# parameters
-n_companies <- 20
-total_exposure_lbk <- 100000000
-total_credit_limit_lbk <- 500000000
-currency_lbk <- "USD"
-
-# define parameters for sampling----
-
-# styler: off
-sector_shares <- tibble::tribble(
-  ~sector,    ~share,
-  "aviation",   0.02,
-  "automotive",  0.2,
-  "cement",     0.02,
-  "coal",       0.03,
-  "hdv",           0,
-  "oil and gas", 0.1,
-  "power",       0.6,
-  "shipping",      0,
-  "steel",      0.03
-)
-# styler: on
-
-# abcd_demo must be sourced first, as the loanbook is created based on this data
-abcd <- r2dii.data::abcd_demo
-
-# replace potential NA values with 0 in production
-abcd_no_na <- abcd %>%
-  dplyr::mutate(
-    production = if_else(is.na(production), 0, production)
-  )
-
-# sample companies based on parameters----
-# seed varies with number of companies and values of sector shares for some randomness
-seed_i <- round(n_companies + sd(sector_shares$share) * 100)
-set.seed(seed_i)
-
-dist_abcd_sector <- abcd %>%
-  dplyr::filter(.data$is_ultimate_owner) %>%
-  dplyr::distinct(
-    .data$company_id,
-    .data$name_company,
-    .data$sector
-  )
-
-n_companies_sector <- sector_shares %>%
-  dplyr::mutate(n = round(.env$n_companies * .data$share, 0)) %>%
-  dplyr::select(-"share")
-
-sectors_to_sample <- sector_shares %>%
-  dplyr::filter(.data$share > 0) %>%
-  dplyr::distinct(.data$sector) %>%
-  dplyr::pull()
-
-sample_sectors <- NULL
-
-for (i in sectors_to_sample) {
-  n_companies_sector_i <- n_companies_sector %>%
-    dplyr::filter(.data$sector == i) %>%
-    dplyr::distinct(.data$n) %>%
-    dplyr::pull()
-
-  sample_sector_i <- dist_abcd_sector %>%
-    dplyr::filter(.data$sector == i) %>%
-    dplyr::slice_sample(n = n_companies_sector_i)
-
-  sample_sectors <- sample_sectors %>%
-    dplyr::bind_rows(sample_sector_i)
-}
-
-# set up basic loan book structure with sampled companies----
-abcd_sample <- abcd %>%
-  dplyr::inner_join(
-    sample_sectors,
-    by = c("company_id", "name_company", "sector")
-  )
-
-sample_sector_codes <- r2dii.data::nace_classification %>%
-  dplyr::filter(.data$borderline == FALSE) %>%
-  dplyr::group_by(.data$sector) %>%
-  dplyr::slice_max(.data$code_level, n = 1) %>%
-  dplyr::slice_head(n = 1) %>%
-  dplyr::ungroup() %>%
-  dplyr::select("sector", "code") %>%
-  dplyr::mutate(
-    sector_classification_system = "NACE",
-    sector_classification_input_type = "Code"
+library(usethis)
+
+source(file.path("data-raw", "utils.R"))
+
+# Source: https://github.com/2DegreesInvesting/r2dii.dataraw/pull/4
+path <- file.path("data-raw", "loanbook_demo.csv")
+loanbook_demo <- readr::read_csv(
+  path,
+  col_types = readr::cols(
+    id_loan = "c",
+    id_direct_loantaker = "c",
+    name_direct_loantaker = "c",
+    id_intermediate_parent_1 = "c",
+    name_intermediate_parent_1 = "c",
+    id_ultimate_parent = "c",
+    name_ultimate_parent = "c",
+    loan_size_outstanding = "d",
+    loan_size_outstanding_currency = "c",
+    loan_size_credit_limit = "d",
+    loan_size_credit_limit_currency = "c",
+    sector_classification_system = "c",
+    sector_classification_input_type = "c",
+    sector_classification_direct_loantaker = "d",
+    fi_type = "c",
+    flag_project_finance_loan = "c",
+    name_project = "c",
+    lei_direct_loantaker = "c",
+    isin_direct_loantaker = "c"
   )
+)
 
-abcd_sample_sector_codes <- abcd_sample %>%
-  dplyr::inner_join(
-    sample_sector_codes,
-    by = "sector"
-  )
+loanbook_demo$loan_size_outstanding <- as.double(
+  loanbook_demo$loan_size_outstanding
+)
 
-loanbook_sample_prep <- abcd_sample_sector_codes %>%
-  dplyr::distinct(
-    .data$company_id,
-    .data$name_company,
-    .data$lei,
-    .data$sector_classification_system,
-    .data$sector_classification_input_type,
-    .data$code
+loanbook_demo <- loanbook_demo %>%
+  mutate(
+    name_direct_loantaker = case_when(
+      name_direct_loantaker == "Holcim Hüttenzement" ~ "Holcim Huttenzement",
+      TRUE ~ name_direct_loantaker
+    ),
+    name_ultimate_parent = case_when(
+      name_ultimate_parent == "Sa Tudela Veguín" ~ "Sa Tudela Veguin",
+      name_ultimate_parent == "Chongyang Hui’An Cement" ~ "Chongyang Hui'An Cement",
+      TRUE ~ name_ultimate_parent
+    )
   )
 
-# TODO: how to handle group_id?
-loanbook_sample <- tibble::tibble(
-  id_direct_loantaker = paste0("C", loanbook_sample_prep$company_id),
-  name_direct_loantaker = loanbook_sample_prep$name_company,
-  id_intermediate_parent_1 = NA_character_,
-  name_intermediate_parent_1 = NA_character_,
-  id_ultimate_parent = paste0("UP", loanbook_sample_prep$company_id),
-  name_ultimate_parent = loanbook_sample_prep$name_company,
-  loan_size_outstanding = NA_real_,
-  loan_size_outstanding_currency = .env$currency_lbk,
-  loan_size_credit_limit = NA_real_,
-  loan_size_credit_limit_currency = .env$currency_lbk,
-  sector_classification_system = loanbook_sample_prep$sector_classification_system,
-  sector_classification_input_type = loanbook_sample_prep$sector_classification_input_type,
-  sector_classification_direct_loantaker = as.numeric(loanbook_sample_prep$code),
-  fi_type = "Loan",
-  flag_project_finance_loan = "No",
-  name_project = NA_character_,
-  lei_direct_loantaker = loanbook_sample_prep$lei,
-  isin_direct_loantaker = NA_character_
-) %>%
-  tibble::rowid_to_column() %>%
-  dplyr::mutate(id_loan = paste0("L", .data$rowid)) %>%
-  dplyr::select(-"rowid")
-
-# add randomly sampled loan size----
-loanbook_demo <- loanbook_sample %>%
-  dplyr::mutate(
-    random_beta = stats::rbeta(n = nrow(loanbook_sample), shape1 = 1, shape2 = 3),
-    share_exposure = .data$random_beta / sum(.data$random_beta, na.rm = TRUE)
-  ) %>%
-  dplyr::mutate(
-    loan_size_outstanding = round(.data$share_exposure * .env$total_exposure_lbk, 0),
-    loan_size_credit_limit = round(.data$share_exposure * .env$total_credit_limit_lbk, 0)
-  ) %>%
-  dplyr::select(-dplyr::all_of(c("random_beta", "share_exposure")))
-
 usethis::use_data(loanbook_demo, overwrite = TRUE)